5.1.1 GPS Satellite Surveying: Some Considerations

GPS SURVEYING versus GPS NAVIGATION

In this section several important matters concerning the GPS satellite surveying methodology will be discussed:

The fundamental differences between the GPS Navigation and GPS Surveying modes of positioning.

The factors which have encouraged the adoption of GPS, in place of competing positioning technologies, for a range of possible land surveying users.

 

The distinction between GPS Surveying and GPS Navigation can be made according to a variety of criteria, for example:

• According to "when", "where" and "how" the GPS technology is applied. This focuses on the applications, and the following simplistic distinction is therefore made: GPS Navigation supports the safe passage of a vessel or aircraft, from the port of departure, while underway and to its point of arrival; while GPS Surveying is mostly associated with the traditional functions of establishing geodetic control, supporting engineering construction, cadastral surveys and map making.
• According to operational aspects, such as the real-time, absolute positioning aspects of Navigation, as opposed to the post-processed, "unhurried", relative positioning characteristics of GPS Surveying.
• According to the type of measurement made and the GPS instrumentation used. GPS Navigation-type receivers are comparatively low-cost, code-correlating instruments that only measure pseudo-range, whereas GPS Surveying receivers are expensive, phase measuring instruments that include many special features and complex software in order to support their function.
• According to the mathematical models used. For example, because the primary measurement in GPS Navigation is the pseudo-range, the biases are dealt with in a more "casual" fashion (with the exception of the clock errors, they are all ignored!). In contrast, GPS Surveying requires a more careful treatment of the biases during the data processing.

The particular characteristics of GPS (land) surveying, the modes of positioning and a review of the treatment of GPS biases (and the resulting operational requirements) are summarised below.

Characteristics of GPS Surveying and GPS Navigation

Some characteristics of GPS Satellite Surveying are:

The points being coordinated are stationary.

GPS data are collected over some "observation session".

Relative positioning modes of operation, and hence high accuracies.

The measurements are made on the L-band carrier wave, hence requiring special instrumentation and software.

Mostly associated with the traditional surveying and mapping functions.

Some characteristics of GPS Satellite Navigation are:

The points being coordinated are generally in motion.

GPS is collected for an "instant", and the solution is obtained in real-time.

Absolute and relative positioning modes of operation, of comparatively low accuracy.

The measurements are typically made on the PRN codes, and requires the processing of pseudo-range data.

Mostly associated with defining safe passage of ships and aircraft.

GPS Positioning Modes

The main positioning modes for GPS surveying and navigation are (section 2.4.1):

• ABSOLUTE or POINT positioning: coordinates are in relation to a well-defined global reference system.
  
• DIFFERENTIAL or RELATIVE positioning: coordinates are in relation to some other fixed point. In GPS surveying this is referred to as baseline determination.
  
• STATIC positioning: coordination of stationary points, either in absolute or relative mode. This is generally synonymous with the SURVEYING mode of positioning, based on the analysis of carrier phase observations.
  
• KINEMATIC positioning: coordination of moving points, either in absolute or relative mode. This is generally the NAVIGATION mode of positioning, based on pseudo-range observations.

As all GPS observations are plagued with biases, hence for both navigation and surveying applications an appropriate combination of measurement and processing strategies must be used to minimise their effect on the positioning results.

There are some distinctions to be made in the way the data is processed in order to minimise the effect of biases in the measurements.

Pseudo-range data is relatively "noisy", and the significant biases are accounted for in the following way:

• In the point positioning mode, satellite clock error is ignored, as it is assumed (after applying the broadcast clock error model -- section 3.3.2) to be smaller than the measurement noise,
• Receiver clock error is estimated in real-time through redundant measurements, because all data is contaminated by the same bias.
• In the relative positioning mode, all satellite and propagation biases are significantly reduced.

This is the NAVIGATION mode of positioning, as results are obtained in real-time (when four or more pseudo-ranges are processed simultaneously). Relative navigation is of higher accuracy asthe primary biases due to orbit error, atmospheric refraction and SA are minimised.

Integrated carrier beat phase data is very precise, hence any contamination by systematic errors is of greater concern than in the case of pseudo-range measurements. Appropriate processing techniques must therefore be used. However, the primary drawback of this data type is its range "ambiguity". In GPS surveying the major biases are accounted for in the following ways:

• Differencing data collected simultaneously from two or more GPS receivers, to several GPS satellites, between satellites and between receivers. This eliminates, or significantly reduces, most of the biases. All position results are therefore expressed relative to (fixed) datum stations.
• The "ambiguity" bias is often estimated, though a weaker solution can be obtained from the appropriate triple-differenced observable (section 6.3).

This is the SURVEYING mode. The fact that the receivers are stationary, and that data is collected over some observation period, permits the ambiguities to be reliably estimated and a strong solution obtained. There are alternative means of estimating ambiguities that permit real-time kinematic baseline determination to be carried out as well.

The GPS Satellite Surveying Methodology: Some Comments

Comments to the operational aspects of GPS Surveying:

• Survey planning considerations are derived from:
  
  • The nature and aim of the survey project --> as for conventional surveys.
  • The unique characteristics of GPS, and in particular no requirement for station intervisibility --> a simplification in survey design.
  • The number of points to be surveyed, the resources at the surveyor's disposal, and the strategy to be used for propagating the survey --> a logistical problem.
  • Prudent survey practice, requiring redundancy and check measurements to be incorporated into the network design.
    
• The requirement for training in the operation of GPS survey receiver hardware, and post-processing software, as well as being aware of calibration and test procedures.
  
• Field operations are characterised by requirements for:
  
  • Setup of antennas over predefined ground marks.
  • Simultaneous operation of two or more GPS receivers.
  • Coordinate data gathering operation so that data collected has the same time-tags, involves the same satellites, etc.
  • Common data collection over some observation session.
  • Coordinated demount of GPS antennas and transport to new stations.
    
• Field validation of data collected, in order to:
  
  • Verify sufficient common data collected at all sites operating simultaneously.
  • Verify quality of data to ensure that acceptable results will be obtained.
  • Where data dropout is high or a station has not collected sufficient data, reoccupation may be necessary.
    
• Office calculations:
  
  • To obtain GPS solutions for single sessions or baselines.
  • To combine the results of single sessions into a network solution.
  • To incorporate external information (for example, local control station coordinates), and hence modify the GPS-only network solution.
  • To transform the GPS results to the local geodetic datum, and to derive orthometric heights.
  • To verify the accuracy and reliability of the GPS survey.

Comments to the GPS Survey Solution:

 

• The fundamental unit of a GPS solution is a 3-D baseline vector joining the antennas of two GPS receivers that have been tracking simultaneously the same satellites. GPS software to carry out the solution task is usually provided by the instrument manufacturer.
• One end of the baseline is held "fixed" (its coordinates are assumed known), and the other station's coordinates are determined relative to it (in effect, the baseline components are estimated).
• Solutions may be obtained from ambiguity-free or ambiguity-fixed double-differenced data solutions, with different resultant accuracies and reliabilities.
• All results are obtained in the quasi-WGS84 reference system, but relative to a fixed station (the WGS84 coordinates of one end of the baseline are assumed known).
• All results refer to the antenna phase centres, and the height of antenna and any offsets must be applied in order to reduce the coordinates to the ground marks. The quality of the baseline vector solution is dependent on, amongst other things:
  • length of the (common) observing session,
  • the number of satellites tracked by the receivers,
  • the quality of the data (multipath and cycle slips, single or dual-frequency data, presence of noise and other biases),
  • the type of baseline solution: triple-difference, double-difference, etc., and
  • the software used to reduce the data.
• If during an observation session more then two receivers were deployed, independent baselines need to be processed, either in a single combined solution or separately.
• If a network needs to be surveyed over a number of sessions (because the number of points is greater than the number of GPS receivers), a combination of separate baseline solutions is needed in a subsequent "network adjustment" step.
• This network solution may then be constrained and transformed into the local geodetic datum if sufficient geodetic control stations (with coordinates in the local datum) are also surveyed as well.
• GPS survey results should be "quality controlled" at all stages of survey and data processing.

Back To Chapter 5 Contents / Next Topic

© Chris Rizos, SNAP-UNSW, 1999